Improving Pseudorandom Generator on Cellular Automata with Bent Functions

Nowadays the practice of researching pseudorandom number generators (PRNG) becomes more scalable because of its spreading in many spheres of computer science and, especially, cybersecurity. The problem is that existing generators are still have many disadvantages in terms of velocity, complexity or flexibility. Thus, the area of researching new algorithms of generating pseudorandom sequences is more than just applicable method, but the target for multiplying cybersecurity from the hardware to application level. This leads to make the set of available and useful PRNG larger and better by their features, like velocity, performance, simplicity in realization. These features match PRNG, based on cellular automata (CA), but not all rules, used in CA are appropriate for their transition functions. Bent functions are perfectly complement statistical weakness of some rules because of their non-linearity without loss of other features.